vector database

Qdrant

Qdrant is a high-performance vector similarity search engine and database built in Rust for generative AI and semantic search.

What the product actually does

Qdrant stores high-dimensional vector embeddings generated by machine learning models and performs approximate nearest neighbor (ANN) search at scale.

It uniquely supports payload-based payload filtering during the vector search process, ensuring exact query results without post-filtering performance degradation.

How pricing works once you are inside

Available as open source for self-hosting. Qdrant Cloud charges based on cluster node RAM, CPU, and disk storage.

Where it breaks down

Vector indexes require substantial RAM allocations; poorly tuned quantization parameters can drive up cloud hosting costs.

What switching costs

Moderate. Raw vectors can be exported to standard parquet/JSON files for import into Milvus, Pinecone, or pgvector.

Sources

Qdrant vector engine documentation and benchmark publications, October 2026.